1. Forecast system version

System name: CMCC-SPS3

First operational forecast run: 1st April, 2018

2. Configuration of the forecast model

Is it a coupled modelYES

Coupling frequency: 
Atmosphere-Ocean: 90 minutes (every third timestep of atmospheric model)

Atmosphere-Land: 30 minutes (also timestep of atmospheric model)

Atmosphere-Sea Ice: 30 minutes (also timestep of atmospheric model)

2.1 Atmosphere and land surface


CAM 5.3 (Atmosphere)

CLM 4.5 (land surface)

Horizontal resolution and grid1° lat-lon approx
Atmosphere vertical resolution46 vertical levels
Top of atmosphere0.2 hPa (60 km approx.)
Soil levels10
Time step30 minutes

Detailed documentation:

CAM Model documentation http://www.cesm.ucar.edu/models/cesm1.2/cam/

CLM Model documentation http://www.cesm.ucar.edu/models/cesm1.2/clm/

2.2 Ocean and cryosphere

Ocean model

NEMO v3.4

Horizontal resolution1/4°
Vertical resolution50 vertical levels
Time step18 minutes
Sea ice modelCICE 4.0
Sea ice model resolution1/4°
Sea ice model levels1 thickness only
Wave modelNO
Wave model resolutionN/A

Detailed documentation: goto NEMO documentation

3. Initialization and initial condition (IC) perturbations

3.1 Atmosphere and land

Atmosphere initialization
ERA InterimECMWF IFS operational analysis
Atmosphere IC perturbations1010

Land Initialization

Forced (obs. atmosphere) monthly run

initialized from 10-year spin-up

Forced (obs. atmosphere) monthly run

initialized from 10-year spin-up

Land IC perturbations33
Soil moisture initializationFrom land initializationFrom land initialization
Snow initializationFrom land initializationFrom land initialization
Unperturbed control forecast?NONO

Detailed documentation:

For more DA details see ECMWF operational analysis documentation at: https://www.ecmwf.int/en/elibrary/16666-part-ii-data-assimilation 

Data assimilation method for control analysis: ECMWF operational

Horizontal resolution of perturbation: N/A

Perturbations in +/- pairs: NO

3.2 Ocean and cryosphere

Ocean initializationC-GLORS
Global Ocean 3D-VAR
Global Ocean 3D-VAR
Ocean IC perturbations48
Unperturbed control forecast?NONO

Detailed documentation:

More ocean data assimilation details available at  http://c-glors.cmcc.it/index/index.html

4. Model uncertainties perturbations:

Model dynamics perturbationsNO
Model physics perturbationsNO

If there is a control forecast, is it perturbed?


Detailed documentation: N/A

5. Forecast system and hindcasts

Forecast frequencyMonthly
Forecast ensemble size50
Hindcast years1993-2016
Hindcast ensemble size40
On-the-fly or static hindcast set?STATIC

6. Other relevant information

The 10 atmospheric perturbed ICs, the 3 land perturbed ICs and the 8 (4 in hindcast mode) are combined to yield 240 (120 in hindcast mode) possible perturbed ICs among which the 50 ICs (40 in hindcast mode) to produce the forecast ensemble are chosen at random.

7. Where to find more information

Sanna, A., A. Borrelli, P. Athanasiadis, S. Materia, A. Storto, S. Tibaldi, S. Gualdi, 2017: CMCC-SPS3: CMCC-SPS3: The CMCC Seasonal Prediction System 3. Centro Euro-Mediterraneo sui Cambiamenti Climatici. CMCC Tech. Rep. RP0285, 61pp. Available at address: